Watch the data take shape.
Two hands-on visuals for learning how behavior becomes data. Respond and watch a cumulative record draw itself, then record a session three ways and see how much each interval method misses.
Visual 1
Cumulative recorder
Every response steps the line up by one. Time keeps moving whether or not you respond, so the line runs flat when you stop and climbs faster when you speed up. The slope is the rate.
Press Respond to start your record.
- Responses
- 0
- Reinforcers
- 0
- Overall rate
- 0 per min
- Last 10 s
- 0 per min
- Time
- 0:00
Reading the record
- Flat line: no responding. A cumulative record can never go down.
- Steep line: a high rate. Compare the slope with the fan of guide lines at the right edge.
- Slash marks: each one is a reinforcer, as on Skinner’s original paper records.
- Dashed line: a schedule change. Its label sits above the graph.
Try this
- Respond steadily for 20 s, stop for 10 s, then respond as fast as you can. Find each part on both graphs.
- Switch to FR 5, then FI 15 s. Do you pause after each reinforcer? Do you speed up as the 15 s run out?
- Switch to Extinction and keep going until you quit. Watch the line bend flat.
- Press Show an example for the classic fixed-ratio, fixed-interval and extinction patterns.
Visual 2
Partial interval, whole interval and momentary time sampling
The top row is what really happened in a 60-second session. The three rows below score the same session with each interval method. Compare each method’s estimate with the true percentage of time the behavior occurred.
Or drag along the top row to draw a bout. Click a bout to remove it.
Estimate compared with what really happened
How each method scores an interval
- Partial interval: scored if the behavior happens at any point in the interval. It tends to overestimate how much of the session the behavior filled.
- Whole interval: scored only if the behavior lasts the entire interval. It tends to underestimate.
- Momentary time sampling: scored if the behavior is happening at the moment the interval ends. It can miss in either direction, and it misses short bouts most.
Try this
- With Long bouts, change the interval from 30 s to 5 s. Watch every estimate move toward the true value.
- With Short, frequent bouts, see how partial interval nearly fills every interval while whole interval scores almost nothing.
- Record live: hold the button during the session and watch partial interval score the instant you start, and whole interval fail the instant you stop.